Optimization Algorithm for Relay Position and Delay of Unmanned Aerial Vehicle For New Type Power Systems

This paper proposes a delay and location optimization algorithm for relay Unmanned Aerial Vehicle (UAV) to address the issue of high queuing delay in the New Type Power Systems. By optimizing the scheduling of UAV positions, the data reception and transmission pressure of public relay UAVs can be alleviated, and queuing delays in the network can be reduced. Firstly, the selection of scheduling UAVs is based on the signal-to-noise ratio and propagation delay of the communication link between UAVs. Then, the position optimization problem of scheduling UAVs is modeled as a Markov decision process. Based on the signal-to-noise ratio and propagation delay of relay UAVs and the queuing delay of public relay UAVs, a reward function is designed, and the Double Deep Q-Network (DDQN) algorithm is used to optimize the position of scheduling UAVs. The simulation results show that compared with existing methods, the algorithm proposed can significantly reduce queuing delay and transmission delay by 18% compared to existing methods, and control the packet loss rate below 8%.